Non-ribosomal peptide synthetases (NRPSs) produce peptides with antibiotic and anticancer activities and are therefore a target for combinatorial or genetic engineering to create catalysts that could generate novel peptides. NRPSs are modular proteins that contain multiple catalytic domains expressed as a single polypeptide. During synthesis, the nascent peptide is transferred from one catalytic domain to the next for further elongation or chemical modification. We have determined the X-ray crystal structures of two adenylate-forming enzymes that suggest that, at different steps of the reaction, the orientation of the C-terminal domain differs by 150 degrees. We have proposed that the closely-related NRPS adenylation domains, which activate the amino acid building blocks and covalently attach them to a second NRPS carrier protein domain, also adopt these two conformations. The magnitude of, and the manner in which these enzymes use, this change is striking and suggests that efforts to engineer the NRPS enzymes to make novel pharmaceuticals will require that steps are taken to avoid steric clashes that arise from the rotation of downstream domains. This domain alternation hypothesis will be investigated through x-ray crystallographic and biochemical analyses of three adenylate-forming enzymes, including a three-domain NRPS. Specifically, we will determine the structures a) acetyl-CoA synthetase, b) an aryl-CoA synthetase, c) a three-domain NRPS protein of which we have expressed a truncated two-domain adenylation domain-carrier protein domain fragment in an active form, and d) a two-domain NRPS protein, which we have crystallized, that serves as the amino acid acceptor for the adenylation domain. Through our structural work and biochemical analyses we will gain insight into the catalytic mechanism of these important NRPS domains, providing the structural foundation for efforts to engineer these catalysts for the development of new drugs.